FORMATION OF THE GERMINAL LAYERS IN TELEOSTEI. 203 



into a number of small spheres, and then by a rupture of its wall set the new- 

 products free, must surely possess more vital properties than are exhibited by 

 the yolk of fish ova. Again, the appearance of the yolk granules is different 

 from that of the yolk spherules. The former are dense and highly refractive, 

 while the latter are only slightly refractive. The " yolk granules " disappear 

 when the egg is fertilised. They do not appear to be incorporated within 

 the protoplasm as so much food, but the dense droplets spread out into a 

 thin film, which becomes indistinguishable amongst the germinal protoplasm. 

 Such dense droplets are found in the ova of the Gadidse and other fishes, and 

 here play a similar part. They are probably oily in nature, and may be pro- 

 duced by the mechanical collection of the oily material contained in the yolk 

 mass. Returning now to Kupffer's statements concerning the structure of the 

 ripe unfertilised ovum, I hold that he is correct in his opinion that the germinal 

 disc, in the strict sense of the word, does not yet exist in the unfertilised ovum. 

 Undoubtedly he was mistaken in concluding that in the unfertilised ovum there 

 is no division into formative and nutritive yolk. This has already been pointed 

 out by Hoffmann. This author, however, appears to regard the germinal 

 protoplasm, which is distributed throughout the yolk in the ripe unfertilised 

 ovum, as homologous with the germinal disc in other fish ova. This is not the 

 case. The germinal protoplasm already existing in the unfertilised ovum does 

 indeed, after fertilisation, become in great part used up in forming the germinal 

 disc, but not entirely so. This protoplasm also increases in bulk considerably 

 at the expense of the yolk before segmentation commences. 



Unfortunately, I have not been enabled to make a thorough study of the 

 development of the ovarian ovum of the herring. I have, however, examined 

 sections of ovaries in various stages of development. The earliest stage which 

 I have observed is one in which the primitive ovum is already surrounded by a 

 follicle. The cell protoplasm contains very little food material, and the nucleus 

 is very distinct. As the ovum increases in size yolk material becomes collected 

 within the egg contents to such an extent as to hide the nucleus and its sur- 

 rounding protoplasm, when only the whole egg is examined. As development 

 goes on it can be made out from sections that the protoplasm increases in bulk 

 at the expense of the yolk. It also spreads out amongst the yolk spheres. At 

 an early stage the cell protoplasm is star-shaped, having a central somewhat 

 thin area in which the nucleus is situated, and around this are a number of 

 radiating protoplasmic processes, which are pushed in amongst the continually 

 increasing bulk of yolk spheres. In this manner the protoplasm comes to form a 

 network supporting the yolk spheres. At a comparatively early stage the ger- 

 minal spot becomes obliterated, and the germinal vesicle also loses its primi- 

 tively well-marked character. In the ripe unfertilised ovum I have not been able 

 to recognise the germinal vesicle. The most approved methods of staining do 



